Beyond Biological Design Artificial Intelligence, Synthetic Novelty and the Convergence of Biosecurity and Cognitive Security

Artificial intelligence now contributes to the design of protein sequences, predicted molecular structures, candidate materials and whole viral genomes, including designs that differ from known natural examples. These capabilities support medicine, materials science and environmental technology. They also raise questions about biological uncertainty, environmental exposure, dual-use capability and the reliability of the information through which emerging risks are assessed. This paper examines the relationship between AI-enabled scientific novelty, biological security and cognitive security. It argues that AI-generated design produces two forms of uncertainty that can interact: physical uncertainty about what a novel design does when it encounters organisms and environments, and epistemic uncertainty about the provenance, interpretation and reliability of information describing it. The analysis draws on AI-assisted materials discovery, protein design, generative genome design and biological watermarking, including experimental evidence that watermarked AI-designed protein binders retained target-binding activity under the conditions tested. It identifies a provenance-to-judgement gap between the existence of a technical provenance signal and a justified decision by a person or institution about what to believe. It proposes a preliminary Biological–Cognitive Security Convergence framework linking design, physical validation, evidential integrity, human judgement and governance, and sets out five propositions for empirical testing. The contribution is conceptual. The paper does not assume that AI-generated substances are inherently hazardous, does not estimate incident probabilities, and does not claim that combined biological and informational incidents have been observed at scale.

Authors

Publication Details

Journal
Zenodo (CERN European Organization for Nuclear Research)
Published
2026-11-14
DOI
https://doi.org/10.5281/zenodo.24220046
Primary Topic
Law, AI, and Intellectual Property
Type
preprint
Controls
|||
ALL TIME
JAN
FEB
MAR
APR
MAY
JUN
JUL
AUG
SEP
OCT
NOV
preprint

Beyond Biological Design Artificial Intelligence, Synthetic Novelty and the Convergence of Biosecurity and Cognitive Security

Sarah Jane Gardner
Zenodo (CERN European Organization for Nuclear Research)
Law, AI, and Intellectual Property
preprint

Beyond Biological Design Artificial Intelligence, Synthetic Novelty and the Convergence of Biosecurity and Cognitive Security

Sarah Jane Gardner
preprint en

Abstract

Artificial intelligence now contributes to the design of protein sequences, predicted molecular structures, candidate materials and whole viral genomes, including designs that differ from known natural examples. These capabilities support medicine, materials science and environmental technology. They also raise questions about biological uncertainty, environmental exposure, dual-use capability and the reliability of the information through which emerging risks are assessed. This paper examines the relationship between AI-enabled scientific novelty, biological security and cognitive security. It argues that AI-generated design produces two forms of uncertainty that can interact: physical uncertainty about what a novel design does when it encounters organisms and environments, and epistemic uncertainty about the provenance, interpretation and reliability of information describing it. The analysis draws on AI-assisted materials discovery, protein design, generative genome design and biological watermarking, including experimental evidence that watermarked AI-designed protein binders retained target-binding activity under the conditions tested. It identifies a provenance-to-judgement gap between the existence of a technical provenance signal and a justified decision by a person or institution about what to believe. It proposes a preliminary Biological–Cognitive Security Convergence framework linking design, physical validation, evidential integrity, human judgement and governance, and sets out five propositions for empirical testing. The contribution is conceptual. The paper does not assume that AI-generated substances are inherently hazardous, does not estimate incident probabilities, and does not claim that combined biological and informational incidents have been observed at scale.

Zenodo (CERN European Organization for Nuclear Research)
Law, AI, and Intellectual Property
AI Navigator

Ask Laika to Summarize, Analyze, and Connect papers live on the map.

Summarize Papers & Methodologies

Extract key findings, datasets, and comparative methods across publications.

Benchmark Rankings & Visual Analytics

Rank top research institutions, authors, funders, topics, and journals by Field-Weighted Citation Impact (FWCI) and paper volume with instant charts.

Connect Distant Disciplines

Bridge topological clusters on the map to find hidden collaborative intersections.

Laika access coming soon